The proposed program of research is focused on examining the validity of the arousal model of interpersonal intimacy. That model identifies the conditions under which directional, reactive changes in the intimacy behaviors (interpersonal distance, eye contact, touch, topic intimacy, etc.) can be expected. Specifically, the model proposes that a sufficient change in intimacy produces arousal change, which becomes differentially labeled, depending on the situational influences. Further, it is proposed that negatively labeled arousal mediates compensatory reactions, while positively labeled arousal mediates enhancing or reciprocal reactions. The basic tests of the model will involve a confederate's manipulation of verbal and nonverbal intimacy relative to a naive subject in the context of a discussion task. Positive and negative expectancies for the discussion will be manipulated as a means of directing the labeling of arousal. Physiological arousal and both verbal and nonverbal intimacy reactions will be monitored continuosuly during the course of the discussion, and evaluative reactions to the discussion will be assessed by post-session ratings. Later experiments are designed to investigate potentially important factors moderating the processes described in the model. In particular, degree of acquaintance between interacting individuals and the personality dimensions of self-monitoring will be examined in the context of the discussion task. Additional influences such as the effect of the rate of change in intimacy and the source of attribution for the intimacy change will be tested if time permits.